THE HERRING FAMILY. 417 



the results of these last naturalists, nor indeed those of their 

 predecessors, for it now appears to be an established fact, as 

 already stated, that 'whitebait' are the fry of the herring, witli 

 which may occur a certain proportion of young sprats. Ex- 

 amples of 'winter whitebait' examined by Day included some 

 sprats as large as 4^- inches long, with ' well-developed roe,' 

 whilst some of the young herring were as long as 7| inches. 



Numerous whitebait from the Thames and elsewhere have 

 been subjected to minute examination by well-known naturalists 

 and no individuals presented structural features which could 

 not be reconciled with the view that they are the fry of the 

 herring and sprat 1 . 



Before following the young herring beyond this stage we 

 may glance at the later brood. 



In the accompanying diagram (p. 418) a graphic system of 

 representing the life-history of the herring has been adopted. 

 This system, though to a large extent self-explanatory, has been 

 referred to in the chapter 011 Rate of Growth, and also in ' The 

 Lesser Sand-eel.' In it the course of the young spring-herring 

 can be traced as narrated above, but we may also note the 

 occurrence of another 'crop' of larval herrings in August and 

 September. The earliest of these may be traced through the 

 same journey as the spring larvae, namely from bottom to 

 mid-water and thence to surface, eventually migrating to the 

 littoral region, but amongst the later larvae the migration 

 becomes less marked (see diagram), and the very latest appear 

 never to leave the bottom, but to migrate shorewards at once, 

 without an intermediate journey through the mid-water and 

 surface. It may be that this is an adaptation to the exigency 

 of a rapidly approaching winter, and the less abundant supply 

 of pelagic food in the later part of the year. In the autumn- 

 series the littoral habit is assumed at an earlier stage, at least 

 as regards length, than in the spring-series. These interesting 

 differences may all be reasonably explained as the direct result 

 of the different environment in each case. 



The autumn-larvae differ from their spring-relatives in 



1 Vide the observations of Prof. Ewart and the late Mr James Duncan 

 Matthews. \tli Report Scot. Fish. Board, p. 98. 



M. P. 27 



